Author’s Accepted Manuscript

Disaster Risk Indicators In Brazil: A Proposal

Based On The World Risk Index

Lutiane Queiroz de Almeida, Torsten Welle, Jörn

Birkmann

www.elsevier.com/locate/ijdr

PII: S2212-4209(16)30064-4
DOI: http://dx.doi.org/10.1016/j.ijdrr.2016.04.007
Reference: IJDRR342
To appear in: International Journal of Disaster Risk Reduction
Received date: 12 February 2016
Revised date: 15 April 2016
Accepted date: 17 April 2016
Cite this article as: Lutiane Queiroz de Almeida, Torsten Welle and Jörn
Birkmann, Disaster Risk Indicators In Brazil: A Proposal Based On The World
Risk Index, International Journal of Disaster Risk Reduction,
http://dx.doi.org/10.1016/j.ijdrr.2016.04.007
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
‘

DISASTER RISK INDICATORS IN BRAZIL: A PROPOSAL BASED ON THE

WORLD RISK INDEX

Lutiane Queiroz de Almeida (Corresponding author)

Prof. Dr.
Universidade Federal do Rio Grande do Norte
Geography Department
Campus Universitário, km 1, BR 101
Lagoa Nova, Natal - RN
CEP 59.078-970
Telefone 55 84 9 99031863
Brazil
lutianealmeida@hotmail.com

Torsten Welle
Dr. Academic assistant
Universität Stuttgart
Institute of Spatial and Regional Planning
Pfaffenwaldring 7
70569 Stuttgart
Deutschland
torsten.welle@ireus.uni-stuttgart.de

JörnBirkmann
Prof. Dr.
Universität Stuttgart
Institute of Spatial and Regional Planning
Pfaffenwaldring 7
70569 Stuttgart
Deutschland
joern.birkmann@ireus.uni-stuttgart.de

Abstract
The DRIB Index- Disaster Risk Indicators in Brazil - provides a tool to help assess, visualise and
communicate different levels of exposure, vulnerability and risk in Brazil. The index may
sensitise public and political decision-makers towards the important topic of disaster risk and
climate change adaptation. This article aims to explore the feasibility and usefulness of such a
national risk index that considers both natural hazard phenomena and social vulnerability. The
exposure to natural hazards was assessed by using four indicators that describe the exposure
of people towards landslides, floods, droughts and sea level rise. Whereas vulnerability
dimension consists of susceptibility, coping capacity and adaptive capacity was calculated on
the basis of 32 indicators which comprise social, economic and environmental conditions of a
society. The county comparison provides an initial ranking of exposure and vulnerability.
Specific analysis of coping and adaptation capacities also indicates that risk or vulnerability are
not pre-defined conditions, but rather are constructed by societies exposed to natural hazards.
‘

The results of the DRIB Index were mapped and classified by means of a GIS system to show
different patterns of exposure, vulnerability and risk on global scale. The national perspective
of risk clearly shows that the vulnerability of a society or a country is not the same as exposure
to natural hazards.The information provided by the DRIB Index highlights the need for
preventive measures towards Disaster Risk Reduction and Climate Change Adaptation in the
country as a whole, but also at regional and local scales. The results showed that the risk is
strongly interwoven with social-economic and cultural conditions and normal everyday life, as
well as with the performance of state institutions dealing with Disaster Risk Reduction (DRR),
in other words, vulnerability. Spatial trends of disaster risk and vulnerability, products of this
research, also have stressed the serious social inequalities between and within regions of the
country, which result in barriers to the development of Disaster Risk Reduction (DRR)in Brazil
as a whole.

Keywords: risk assessment; risk index; natural hazards; vulnerability; Brazil.

1. Introduction

The year 2015 is the 25th year since the creation of the International
Decade for Natural Disaster Reduction (IDNDR) and the International Strategy for Disaster Risk
Reduction (ISDR) proposed by the United Nations. This year also witnessed the Third UN World
Conference on Disaster Risk Reduction, where the Sendai Framework for Disaster Risk
Reduction 2015-2030, was put into place, reviewing the implementation of the Hyogo
Framework for Action 2005-2015.
In this historic year, despite significant improvements in reducing disaster
risk at the local, national, regional and global levels as indicated in various reports prepared by
organisations linked to the United Nations, the damages caused by natural hazards also have
increased, especially those caused by disasters1. Even with all the knowledge and the recent
evolution in terms of mapping, risk assessment, satellites, early warning, geographic
information system (GIS), remote sensing, constructiontechniques, etc., disasters still occur
inNepalin April2015, where over 8,800 people lost their lives and more than 23,000 injured
people (Barry, 2015). It is more important than ever to develop risk and vulnerability
assessments and frameworks, contributing to understanding the risks of natural hazards and
the potential impacts of climate change on human society, at all levels (local, national,
regional, global).
According to Welle and Birkmann (2015), 20 years of research in the field of
natural hazards clearly indicates that not only do extreme events lead to disaster, but
especially the conditions of societies exposed to natural hazards determine whether a natural
phenomenon can trigger a disaster or not. Based on the experience of the Hyogo Framework
for Action and the priorities for action defined in Sendai Framework for DRR (2015),
particularly Priority Number 1 (understand disaster risk), this study aims to contribute to
understanding disaster risk in Brazil through the elaboration of indicators and indices to
synthesise and operationalise the abstract concepts of risk and vulnerability within a modular
structure of four key components: exposure to natural hazards; susceptibility as a function of
public infrastructure, housing conditions, economic frameworks; coping capacities as a
function of governance, disaster preparedness, medical services, social and economic security;

1
Third United Nations World Conference on Disaster Risk Reduction Sendai, Japan, 14-18 March 2015, p.
03 – UNISDR, 2015.
‘

and adaptation to future natural events and climate changes (Welle and Birkmann, 2015;
Birkmann et al., 2011).
The DRIB Index project – Disaster Risk Indicators in Brazil aims to serve as a
tool to help assess, visualise and communicate different levels of exposure, vulnerability and
risk in Brazil. Furthermore, the index may sensitise public and political decision-makers
towards the important topic of disaster risk and climate change adaptation (Birkmann et al.,
2011). The DRIB Index aims to explore the feasibility and usefulness of such a national risk
index that considers both natural hazard phenomena and social vulnerability. The county
comparison provides an initial ranking of exposure and vulnerability. In addition, specific
analysis of coping and adaptation capacities also indicates that risk or vulnerability are not pre-
defined conditions, but rather are constructed by societies exposed to natural hazards. The
information provided by the DRIB Index highlights the need for preventive measures towards
Disaster Risk Reduction (DRR) and Climate Change Adaptation (CCA) in the country as a whole,
but also at regional and local scales.
The exposure to natural hazards was assessed by using four indicators that
describe the exposure of people towards landslides, floods, droughts and sea level rise.
Whereas vulnerability dimension consists of susceptibility, coping capacity and adaptive
capacity was calculated on the basis of 32 indicators which comprise social, economic and
environmental conditions of a society. The results of the DRIB Index were mapped and
classified by means of a GIS system to show different patterns of exposure, vulnerability and
risk on global scale. The national perspective of risk clearly shows that the vulnerability of a
society or a country is not the same as exposure to natural hazards.

1.2. Risk and vulnerability indicators: between the potential and the real

In studies of disaster risk (UNISDR, 2009; Wisner et al., 2004;Hewitt, 1997),

it is necessary to make an initial distinction that opposes what is potential (what ‘may’ happen)
and what actually is produced or occurs. There is no agreement between these two
dimensions because the risk (potential) may not materialise in disaster. The risk presents a
probabilistic dimension that a disaster, unfortunately, no longer has (Dauphiné, 2005).
Potential issues involved in risk science are defined by terms such as risk,
danger, fluctuation, random potential damage. On the other hand, to describe a reality, a
phenomenon that truly occurs, terms apply such as catastrophe, disaster, cataclysm, flagellum,
calamity, sinister. That way, there is not ever any concordance, be it temporal, spatial or in
intensity between risk and disaster.
In an evident way, temporally speaking, risk precedes disaster; spatially,
there are real risk zones (relatively large), while disasters are generally more localised.
Operationally, the variable overlap and interactions of the determinants of the risk (hazards,
vulnerability, coping and adaptive capacities) contribute to the production of the risk map.
Damage and loss determine the geographic space of their unfavourable overlap and its
severity, while disasters indicate the places where the interaction is most adverse.
In this context, Hewitt (1997) clarifies that if these hazardous conditions
and relationships (and their suitable measures) can be identified, they provide a framework for
technical analysis and practical response. For this, Hewitt highlights GIS as an important tool
for strategy and planning, monitoring and managing the risks.
Therefore, understanding the interaction between hazards, exposure, and
vulnerability is crucial for effective disaster prevention. Risk assessment is a process that
determines the nature and extent of the risk, through hazard analysis and the evaluation of
existing conditions of vulnerability that together could potentially harm exposed persons,
properties, services, livelihoods, and the environment on which they depend (UNISDR, 2009).
‘

In the context of the latest state-of-the-art approaches to assessing disaster

risk, Birkmann et al. (2011) differentiateatleasttwomajorschoolsofthought: the
DRRcommunityand the CCA2. Within thesetwo communities, several frameworks and
conceptualisations exist for how to assess risk and vulnerability, including quantitative and
qualitative assessment methodologies (Wisner et al., 2004; Birkmann, 2013; International
Federation of the Red Cross and Red Crescent Societies, 2008; IPCC, 2012). In this regard, there
is the systematisation of climate change-related events, vulnerability, exposure, risk, and
development in the SREX report (IPCC, 2012), as examples of the integrated approach to
disaster risks in the context of climate change, reflecting recently established co-operation
between DRR and CCA communities.
Different conceptual frameworks in assessing risk and vulnerability have
historically been developed, with different integration approaches and levels depending on the
schools of thought they belonged to. Birkmann (2013) made an important overview of these
frameworks, highlighting the role of measuring vulnerability in disaster risk assessment.
Another aspect of risk assessment concerns the scale of analysis, mainly
with the production of risk and vulnerability indicators and indices. Birkmann (2006 and 2013)
published an important book about measuring vulnerabilities against natural hazards,
highlighting various frameworks, concepts and operational scales in global, national, sub-
national and local assessment approaches. On a global scale, it emphasises the Global Disaster
Risk analysis, contribution to the GAR 2009 United Nations - Global Assessment Report on
Disaster Risk Reduction (UNISDR, 2009). In this project, the data generated have been made
available online through the PREVIEW Global Risk Data Platform3.
Another important concept of risk and vulnerability assessment developed
by UNU-EHS and the Alliance Development Works is the WorldRiskIndex (WRI). In this
approach, also on a global scale, the aggregate framework of WRI allows its application for risk
and vulnerability assessment at national and local level, and application on a national/local
scale to Indonesia, is described.
Several other risk assessment proposals have already been implemented on
thenational level, such as in the United States (Cutter et al., 2003), Asian countries (ASEAN
Disaster Risk Management Initiative, 2010), and Pacific countries (Pacific Catastrophe Risk
Assessment and Financing Initiative - PCRAFI).
Specifically, there are several examples of assessment for risk and
vulnerability to disasters in Brazil, both in academia and in the technical-administrative sphere
of the country, a few of which will be mentioned4. There are also various studies from the
point of view of the scale of analysis, either local, state and nationwide, as well as on specific
spatial selection, such as metropolitan areas, and, watersheds among others. Similarly, one
can still distinguish between research carried out according to the specific schools of thought,
such as DRR, CCA, political economy, or social ecology. However, there are some aspects that
make risk assessment and vulnerability in Brazil still problematic.

2
There are, nonetheless, other schools of thought which the authors highlight in other publications
(Birkmann, 2013;Birkmannetal., 2013.).
3
http://preview.grid.unep.ch/.
4
There is nointentionto discussthe mainresearch conducted in Brazilonriskandvulnerability assessment,
let alone to discuss theimpressionsof theauthors aboutacademic/technicalproduction on the
subject,withoutintentto exhaust the subject. For a more completeliterature review,
seetheBibliographicDatabaseorganised by the BrazilianResearch Networkfor DisasterRisk Reduction:
https://www.zotero.org/groups/redriscos/items/. The Network, which is in developmentsince2013, is
an initiativeof researchers andresearch institutionsfrom all over Brazil, the National Protection and Civil
Defense Secretariat, and the International Strategy for Disaster Risk Reduction, and other
organisationsarea.
‘

There are few studies that take a holistic approach, integrating exposure,
hazards, vulnerability, and risk into the analysis (Almeida, 2010; 2012). Many studies are
sectorial, that is only analysing specifically the vulnerability, or exposure in isolation, or
hazards assessment. Taking into account the scale or the spatial area, most studies
concentrate on analysing local aspects of disaster risk - household, district, county, river basin
(which is productive), but there are few approaches at the state level.
On the technical-administrative level, risk studies have focused efforts on
disaster recurrence analysis (UFSC.CEPED, 2013), disaster occurrence area mapping, or on
geotechnical mapping aimed at structural measures such as disaster prevention taken under
government in its various spheres. Yet these studies fail to integrate exposure and
vulnerability, considering vulnerability as the factor that determines whether a disaster occurs
or not.
Even considering the territorial concerns of the country with regard to
disaster risk, there is no methodology or framework available to assess risk and vulnerability
for the whole country, although there is enough available data to research on these topics,
providing important information on disasters Brazil. A recent initiative was conducted by the
Institute of Applied Economic Research (IPEA) for mapping the social vulnerability in Brazilian
counties (Costa and Marguti, 2015). Another aspect that should stand out is the enormous
concentration of knowledge production on risk, vulnerability and disaster
preparedness/reduction in Brazil, mainly in the South and Southeast Regions (both in
academia and in the technical-administrative sector). It is important to balance this spatial
difference and foster more research and partnership (academic and other) for knowledge
production on this topic in other parts of the country, especially in the Northeast Region,
which is historically the region that has presented the greatest number of officially reported
disasters by the national civil defence (UFSC.CEPED, 2013) and holds greatest vulnerability
indices, although on the number of deaths, people affected and economic losses by disasters
could be greater in the Southeast and South Regions.
Therefore, the DRIB Index is a proposal for assessing risk and vulnerability
to disasters in Brazil - on the county level throughout the entire country -, which provides a
comparative analysis of the spatial patterns of exposure, vulnerability and risk among counties,
metropolitan areas, states and regions in the country, contributing to the academic/scientific
debate on the issues involved, as well as providing basic information for decision-making
regarding the Disaster Risk Reduction and Management in the country as a whole.
In this regard, DRIB Index can still work with Brazil for what was defined in
Sendai Framework for Disaster Risk Reduction 2015-2030 (UNISDR, 2015, p. 9), in Sendai,
Japan, March 2015, in which it defines as a priority on the national and local scales,
understanding disaster risk.

1.3. Concept of the DRIB Index

The concept of the DRIB Index is based on the WorldRiskIndex (Birkmann et

al., 2011;.Welle et al., 2012;Welleand Birkmann, 2015), whose theoretical concepts and
understanding of risk, within the framework of natural hazards and disaster risk community,
state that disaster risk derives from a combination of physical hazards and vulnerability of
exposed people (UNISDR, 2004; Wisner et al., 2004; Birkmann, 2006; IDEA, 2005; IPCC, 2012).
A broad range of researchers who participated in preparing the IPCC-SREX
agrees that a hazardous event is not the only driver of risk. These researchers are confident
that the level of adverse consequences is largely determined by the vulnerability and exposure
of societies and socio-ecological systems (Cardona et al., 2012).
Also according Cardona et al. (2012), disaster can be understood as the
materialisation of risk and means ‘a becoming real’ of latent conditions in the social
constructions. To understand the socially constructed notion of risk is to understand above all
‘

that disaster risk conditions are significantly determined by vulnerability. In turn, vulnerability
is related to predisposition, susceptibility, fragility, weaknesses, deficiencies or lack of
capacities. These conditions are conducive to adverse effects of disasters upon exposed
groups5. This is a comprehensive conception of risk, accounting for the overlap between, on
the one hand, exposure to natural hazards and, on the other, socio-economic/cultural
conditions and processes, i.e. vulnerability, which is itself composed of susceptibility, coping
capacities and adaptive capacities. For the purposes of this study, it is considered, as in
Birkmann et al. (2010, 2011) and Cardona et al. (2012), that the concepts of coping and
adaptation are different categories and issues, aspects in need of further illumination.
The structure of the DRIB Index is based primarily on the WorldRiskIndex
framework, whose conception is linked to several other integrated and holistic frameworks
that have sought to assess vulnerability and face the challenge of understanding response
capabilities (Bogard and Birkmann, 2004; Cardona, 1999, 2001; Birkmann 2006, 2011a). The
modular structure of the WorldRiskIndex may be applied not only at the national, but also at
the regional and local levels. A local scale assessment was performed by Welle et al. (2011),
specific to Indonesia and focusing only on urban areas at national scale (Welle et al., 2014).
Above all, the DRIB Index aims to capture and measure four major
components (figure 1): exposure to natural hazards; susceptibility of the exposed
communities; coping capacities;adaptive capacities.

Figure 1: Structure of the Index and the Indicator System. Source: Almeida, 2015. The concept
of DRIB Index is based on WorldRiskIndex (Birkmann et al., 2011;.Welle et al., 2012; Welleand
Birkmann, 2015).

In accordance with United Nations Development Programme (UNDP, 2004),

exposure defines risky elements, those artefacts and people who are exposed to a hazard, that
is, elements located in an area within which hazardous events can occur (Cardona, 1990;
UNISDR, 2004, 2009b). Therefore, if a population and its resources are not located in (exposed
to) potentially hazardous spaces, no disaster risk problem exists. Exposure is a necessary but
not exclusive determinant of risk. It is possible to be exposed but not vulnerable. However, to
be vulnerable to an extreme event, it is necessary to also be exposed (Cardona et al., 2013).

5
Towards a more substantial understanding of the concepts ofrisk, vulnerability, hazard, exposure, and
other concepts operationalized in this research, see the following references: UNDRO 1980; Cardona,
1986, 1990, Maskrey, 1993; Cannon, 1994, 2006; Blaikie et al., 1996; Bogardi and Birkmann, 2004;
UNISDR 2004,2009; Birkmann, 2006, Thywissen, 2006.
‘

Within the natural hazard research community, exposure extends to

entities (people, resources, infrastructure, production, assets, services, ecosystems, and
coupled social-ecological systems) exposed or prone to being affected by a hazard event.
Exposure can further be differentiated in spatial (geographical exposure)
and temporal terms. For this study, exposure is operationalised by dividing population exposed
to a certain hazard and total population of the county, which allows the comparison between
the exposed counties.
Meanwhile, there are some specific challenges in measuring exposure that
should be considered. On the one hand, exposure to various natural hazards and creeping
changes, e.g. sea level rise and droughts, differs according to specific characteristics such as
geographic extension, interval of recurrence and magnitude, on the other hand, there is the
case of comparison between large and small counties, imposing difficulties and inaccuracies in
the results and analysis.
Notably, two distinct categories of natural hazards are addressed in the
DRIB Index: sudden-onset hazards (landslides and floods) and slow-onset (creeping) hazards
(droughts and sea level rise).
The DRIB Index clearly prioritises hazards widely distributed in Brazil that
account for serious harm to people and their properties in terms of losses and numbers of
deaths and people affected. For the period 1991-2012, the most frequent and devastating
natural hazards officially reported in Brazil, were droughts, floods and storms, accounting for
91.07% of the total records (UFSC.CEPED, 2013). In terms of number of deaths, focus is placed
on flash floods, landslides, and gradual floods which caused 87.15% of all disaster-related
deaths in Brazil in that period. In addition, emerging hazards in the context of climate change,
in the case of sea level rise, were taken into consideration in the elaboration of the final
index.In this context, the four following natural hazards were investigated in the research:
floods, landslides, droughts, sea level rise.
Susceptibility is another component of the DRIB Index and refers to
exposed communities or other exposed elements (infrastructure, ecosystems etc.) that make
them more likely to experience damage and be adversely affected by a natural hazard or by
climate change. Whereas this component is closely related to structural characteristics such as
infrastructure, economic capacity, and nutrition, it can provide basic evidence of the specific
vulnerabilities of society (Welle and Birkmann, 2015).
Coping capacity is the ability of a group or society, organisation and
systems, using available tools and resources, to face and manage emergencies, disasters, or
adverse conditions that could lead to a harmful process caused by a hazardous phenomenon.
(UNISDR, 2009). The coping capacity of populations affected by natural disasters is a key
concept in vulnerability assessment (Billing and Madengruber, 2005).
It is necessary to emphasise that, for this study, coping and adaptive
capacities are distinct but complementary notions. In the DRIB Index, coping is a direct
response to the impact of a given hazard event, entailing the immediate reaction during a crisis
or disaster, that is, coping is hazard-related and rather short-term oriented.
In light of the short-term orientation of coping capacity, the scientific
community DRR deals with phenomena in the short term, whereas the concept of adaptive
capacity is most often used in the framework of the CCA community, and refers to adaptation
in terms of long term processes, including structural changes (chapter 1 of SREX).
In the DRIB Index, adaptation covers capabilities, measures and strategies
that enable communities to change and transform in order to deal with the negative
consequences expected from natural hazards and climate change. Therefore, these capabilities
focus on features that allow socioeconomic and cultural changes within society.

2. Methodology
‘

This section provides an overview of selected indicators, datasets, as well as

the calculations of the DRIB Index and its specific components. In order to implement the
concept, it was necessary to fulfil certain standards and specific criteria for selecting the
indicators and data: freely available data (in the World Wide Web); practitioners and scientists
monitored and evaluated the selection and characterisation of components, sub-categories
and indicators of the index (in the case of the WorldRiskIndex model particularly);indicators for
exposure allow for specific comparison of very different hazard types; vulnerability indicators
are of a generic nature in order to be relevant to different hazards (multihazard perspective);
indicators are reliable, analytically and statistically sound; reproducible; appropriate in scope;
understandable; easy to interpret and comparable; data are compiled on a regular basis in
order to enable future monitoring of the indicators.
In developing DRIB Index, various methodologies were used, such as
statistical analysis (using Microsoft Excell and IBM SPSS) and spatial analysis using geographic
information system (GIS). For the spatial analysis and mapping, the values of the calculated
indices were divided into five classes, using quantile method, which is integrated into the
ArcGIS 10 software. Hence, each class has an equal number of features and all calculated
indices differ in their value ranges, but qualitative classification of ‘very high – high – medium –
low – very low’ was established.
In this regard, the individual components of exposure and vulnerability are
more relevant to communication and decision-making than the aggregate total index, since an
aggregation always results in a loss of differentiation (see Birkmann et al., 2011).

2.1 Indicators

The DRIB Index was calculated based on 36 indicators, 4 of which refer to

exposure to natural hazards, and 32 pertain to the societal vulnerability. The figure 2
(indicators-synthesis) displays these indicators, their respective components (exposure,
susceptibility, coping capacity, adaptive capacity), and sub-categories. The three sub-
categories – storms, nutrition, and social network, family and self-help – are mentioned
although they could not be integrated into the calculation, but are recognised as important
elements which should be in the concept6.
The indicator selection is based on the WorldRiskIndex perspective and
relates to the eight Millennium Development Goals and the Hyogo Framework for Action of
the United Nations.
The raw data of all selected indicators were extracted from various global
and Brazilian official databases (see descriptions of the indicators in the Supplementary
material). For the index aggregation, all indicators have been transformed in dimensionless
rank level between 0 and 1; that is, they can be read as percentage values for better
comprehension.
The figures that show the modular composition of the indices for
susceptibility, coping capacity, adaptive capacity, including their respective weights can be
seen in the Appendix A. The tables 1, 2 and 3 (transformation methods-indicators) displays the
data source and transformation methods used for the 32 vulnerability indicators. The
indicators for calculating exposure are explained in the next section.

Table 1: Data sources and transformation methods for susceptibility indicators.

6
These sub-categories present lack of data availability, data without validation possibility, or
inappropriate data scale to the present research.
‘
Indicator
A) % people in households with
inadequate water supply and
sanitation
B) Share of population in irregular
clusters (slums)
C) % people in households with
inadequate wall materials
D) Degree of urbanization
E) Dependency ratio
F) % Vulnerabletopoverty
G) Per capita income
H) Gini index
Table 2: Data sources and transformation methods for coping capacities indicators.
Indicator
A) Governmental corruption index
B) Structural measures to reduce
disaster risk
C) Disaster risk management to
Floods
D) Disaster risk management to
Landslides
E) Vulnerable population to
disasters (floods, landslides) is
registered in housing programs
F) Local structure for disaster
response
G) Number of physicians per 1,000
inhabitants
H) Number of hospital beds per
1,000 inhabitants
I) Coverage level of income –
transfer program (Bolsa Família,
2012)
Table 3: Data sources and transformation methods for adaptive capacities indicators.
Indicator
A) Illiteracy rate - 15 years or
more
B) % 15-24 years inprimary
C) % 18-24 years in secondary
D) % 15-17 years in tertiary
E) Institution responsible for
the formulation, coordination
and implementation of policies
for women with specific
budget
F) County has a Plan of Policies
for Women
G) % Of mothers household
Datasource Unit/value Transformationmethod
Human Development Atlas of Brazil 2013 – UNDP, Percentage Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and normalized to 0 to 1
Fundação João Pinheiro
Censo Demográfico 2010 – IBGE (2010) Percentage Logarithm (LN) transformation and
normalized to 0 to 1
Human Development Atlas of Brazil 2013 – UNDP, Percentage Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and normalized to 0 to 1
Fundação João Pinheiro
Brazilian Directory of Geospatial Data Ratio Normalizedto0to 1
Human Development Atlas of Brazil 2013 – UNDP, Number Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Percentage Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Ratio Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Ordinal Logarithm (LN) transformation and
Instituto de Pesquisa Econômica Aplicada and scalefrom0to normalized to 0 to 1
Fundação João Pinheiro 1
Datasource Unit/value Transformationmethod
Boll, 2010 (Master Thesis) Number Logarithm (LN) transformation and
normalized to 0 to 1
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Normalizedto0to 1
IBGE (2014)
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Logarithm (LN) transformation and
IBGE (2014) normalized to 0 to 1
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Logarithm (LN) transformation and
IBGE (2014) normalized to 0 to 1
Profile of Brazilian Municipalities (MUNIC) - 2013 – Ordinal Logarithm (LN) transformation and
IBGE (2014) scalefrom0to normalized to 0 to 1
1
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Logarithm (LN) transformation and
IBGE (2014) normalized to 0 to 1
Brazilian Directory of Geospatial Data Number Logarithm (LN) transformation and
normalized to 0 to 1
Brazilian Directory of Geospatial Data Number Logarithm (LN) transformation and
normalized to 0 to 1
Brazilian Directory of Geospatial Data Percentage Logarithm (LN) transformation and
normalized to 0 to 1
Datasource Unit/value Transformationmethod
Human Development Atlas of Brazil 2013 – UNDP, Ratio Logarithm (LN) transformation
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Ratio Logarithm (LN) transformation
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Ratio Logarithm (LN) transformation
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro
Human Development Atlas of Brazil 2013 – UNDP, Ratio Logarithm (LN) transformation
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro
Profile of Brazilian Municipalities (MUNIC) - 2013 – Valuesbetween0and Normalizedto0to 1
IBGE (2014) 1
Profile of Brazilian Municipalities (MUNIC) - 2013 – Valuesbetween0and Normalizedto0to 1
IBGE (2014) 1
Human Development Atlas of Brazil 2013 – UNDP, Percentage Logarithm (LN) transformation
‘

heads without complete
primary, with children under
15 years
H) Specific policies and actions
for the environment
I) Share of deforestation
J) Conservation areas
L) Fire spots (2014)
M) Legislation and Planning
Instruments
N) Specific planning tools to
prevent disasters
O) Commitments Schedule of
the Millennium Development
Goals - Manager joined the
Agenda of Commitments
P) Life expectancy at birth
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Normalizedto0to 1
IBGE (2014)
Amazon Forest: Brazilian Institute of Space Ratio Logarithm (LN) transformation
Research. and normalized to 0 to 1
Other biomes: Ministry of Environment
Ministry of the Environment Ratio Logarithm (LN) transformation
and normalized to 0 to 1
Burning and Forest Fires - Orbital Monitoring and Number Logarithm (LN) transformation
Fire Risk. INPE, National Institute for Space and normalized to 0 to 1
Research.
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Normalizedto0to 1
IBGE (2014)
Profile of Brazilian Municipalities (MUNIC) - 2013 – Average Normalizedto0to 1
IBGE (2014)
Profile of Brazilian Municipalities (MUNIC) - 2013 – Valuesbetween0and Normalizedto0to 1
IBGE (2014) 1
Human Development Atlas of Brazil 2013 – UNDP, Years Logarithm (LN) transformation
Instituto de Pesquisa Econômica Aplicada and and normalized to 0 to 1
Fundação João Pinheiro

Figure 2: Components with sub-categories and selected indicators for the DRIB Index.
Source: own figure

2.2 Exposure

The selection of natural hazards was based on two aspects: the natural
hazards that occurred more frequently, and that caused more victims (people affected and
deaths) between 1991 and 2012 (UFSC.CEPED, 2013). In this context, three selected natural
hazards - floods, landslides, droughts - produced 85.8% of the disasters reported in Brazil in
that period, accounting for 85.8% of people affected by disasters, and causing 94.72% of
‘

disaster-related deaths. The inclusion of other types of hazards was seriously considered, but
discarded due to lack of suitable data.7Additionally, sea level rise was taken into account, since
it is very likely that due to further climate change, sea level rise will affect many low-lying
coastal zones and delta regions. In 2010, according to the IBGE8 Census, 26.58% of the Brazilian
population lived in cities located in the coastal zone.
Within the DRIB Index, exposure was operationalised as physical
exposure(figure 3), which means the potential annual average number of individuals exposed
to floods, droughts, landslides and sea level rise in Brazil. The PREVIEW Global Risk Data
Platform, a globally available dataset generated by different UN agencies (UNEP, UNDP/BCPR
(GRIP); UNISDR) and the World Bank, was used for the specific cases of exposure to floods,
landslides. The PREVIEW Platform is a multi-agency effort to share spatial data on global risk
related to natural hazards. In this scope, each set of hazard data represents an annual
estimation of the exposed population. This includes a probabilistic component in the
frequency of the respective hazard, and the information on population distribution based on
Landscan TM Population Database (ESRI Grid population, 1 sq resolution for the year 2010).

Figure 3: Technical flowchart for Exposure calculation.

Source: organized by Almeida, 2015.

7
This is the case for storms, which represent as considerable number of events in Brazil, 7.1% of the
total deaths and 7.07% of people affected, mainly in the South Region of the country, where 77.4% of
the nation’s storm events occur (UFSC.CEPED, 2013). For more information on this, see the description
of the indicators in the supplementary material.
8
Brazilian Institute of Geography and Statistics.
‘

The number of people exposed per hazard and per Brazilian county was
derived by calculating the zonal statistics with ArcGIS 10. It should be noted that the global
data for exposure are based on model calculations and thus some uncertainty in the
calculation model has to be taken into account. Considering emerging risks posed by climate
change, as well as the considerable population group that inhabits the coastal area in Brazil, it
was decided, in accordance with the original methodology, to integrate exposure to sea level
rise into the DRIB Index. Since there was no information on physical exposure to sea level rise
available in the PREVIEW Platform, this information was derived from the SRTM 9 image
database from the EarthExplorer website (US Geological Service - USGS). The information was
used to generate a relief contour of 1 meter, corresponding to sea level rise scenario. The area
impacted by the projected sea level rise was then used to determine the exposed population,
based on Brazil’s population grid (mentioned above), combined with zonal statistics created
using the ArcGIS 10. Nevertheless, the indicator for population exposed to sea level rise
measures the population proportion currently (as of 2010) living in an area that may be
affected by 1 meter sea level rise. This means there is a lack in terms of probabilistic
component intrinsic to the other three hazards for estimating the exposed population.
The population exposed per county in Brazil was estimated through the
zonal statistics calculation. However, to reduce the impact of sea level rise exposure on the
overall exposure index, this indicator was weighted10 with 0.5, since it is a gradual process and
lacks a probabilistic component. The same weighting (0.5) was applied to the population
exposed to drought, given that this calculation and the data might overestimate the number of
exposed population, taking into account the complexity of this phenomenon and the less
accurate data (Peduzzi et al., 2009). Finally, the entire population exposed to hazard was
calculated and divided by the total population of each county to obtain a single exposure index
by county.The input data for exposure have been converted into non-dimensional ranks with
values between 0 and 1 and normalized using Logarithm (LN) transformation (IBM SPSS).

2.3 Susceptibility

The susceptibility index provides an overview of the indicators used to

describe susceptibility of societies and social groups to natural hazards, at the local level
(county) with local and regional comparison. This component comprises eight indicators
distributed among four sub-categories: economic capacity and income,poverty and
dependencies, housing conditions and public infrastructure. Nutrition data could not be
integrated into the index aggregation, because data available are not suitable to the scale
analysis of this research.The input data for susceptibility indicators (table 4, figure 2) have
been converted into non-dimensional ranks with values between 0 and 1. Hence, the
susceptibility index is aggregated according to the stated weights (see Appendix A), and their
spatial information contained on the map (figure 7).

2.4 Coping capacity

The coping capacity index was calculated based on several indicators that
determine the capacity of a county to immediately manage or react to the impact of a
hazardous process. This index captures the material conditions and resources used by a society
in an emergency, such as material protection or medical services, as well as structures that
could inhibit the coping of a county, such as corruption, weak governance, and lack of disaster
preparedness. It is necessary to explain that the sub-category social networks, family and self-
help could not be included due to insufficient data on the county level. For the aggregation of

9
Shuttle Radar Topography Mission.
10
The justification of weighting is described in Welle and Birkmann (2015).
‘

the DRIB Index, the lack of coping capacities is included, since the overall sum of the
vulnerability components will be a measure of deficiencies in societal capacities to deal with
natural hazards and climate change impacts. In this respect, the value of each indicator is
subtracted from 1 to compose the lack of coping capacities index, that is displayed in the map
(figure 8).

2.5 Adaptive capacity

Indicators for capturing characteristics of adaptive capacities of a county

and its population seek to demonstrate the long-term response capacities to natural hazards
and/or climate change. This component indicates the ability of a society/community to
transform or adapt in an effort to reduce vulnerability to these changes and impacts. The
adaptive capacity component contains five sub-categories: education and research, gender
equity, environmental status/ecosystem protection, adaptation strategies and investments. As
with the lack of coping capacities index calculation, adaptive capacity was also aggregated into
the overall index referring to lack of adaptive capacities.

2.6 Calculation of the DRIB Index

It was previously demonstrated that each component of the DRIB Index –

exposure, susceptibility, lack of coping capacities and lack of adaptive capacities – has been
calculated separately. Intending to obtain an overview of vulnerability, the components
susceptibility, lack of coping capacities and lack of adaptive capacities were aggregated into
one vulnerability index that characterises societal conditions and processes essential to dealing
with disaster risk in the context of climate change and natural hazards. Overall, the
vulnerability index indicates whether a disaster may ensue if a natural hazard occurs.
Ultimately, the vulnerability index is multiplied by exposure, comprising the magnitude and
frequency of hazards, to obtain the risk.The fundamental hypothesis for the multiplication of
exposure is, that if a vulnerable society is not exposed to a natural hazard, the level of risk will
be zero (even if it is known that in practice there is no such thing as zero risk). The DRIB Index
results have been calculated with non-dimensional ranks with values between 0 and 1. The
figure 3 displays the formula that outlines how the DRIB Index was calculated, including equal
weighting of the components of susceptibility, lack of coping capacities and lack of adaptive
capacities. A detailed list and description of all indicators and data sourcescan be seen at the
supplementary material.
‘

Figure 3: Calculation of the DRIB Index.

Figure 8: Equations used for components calculation.

3. Results
The application of the theoretical concept of risk was carried out at the
national level for comparison between counties. In accordance with data availability, 5,565
counties were taken into account to calculate exposure, susceptibility, lack of coping
capacities, lack of adaptive capacities, the vulnerability index and DRIB Index. All calculated
indices were divided into five classes for the spatial analysis and the development of maps.
Quantile method, which is integrated in the software ArcGIS 10, was used to the group
classification. Hence each class contains an equal number of features. Supported by the
calculated values, each class is also described in a qualitative way such as following
classification schema: very high, high, medium, low, very low.

3.1 Exposure

The map of exposure (figure 6) displays the potential exposure of individual

counties to natural hazards, such as landslides, floods and droughts, as well as the exposure of
‘

populations to one-meter sea level rise for each county on the Brazilian coast (see exposure
maps for each hazard type in the figures 4 and 5). It is clearly demonstrated that the main
hotspot regions for exposure are counties located in the South Region, mainly in the
mountainous regions and floodplains; and North Region, above all in the watershed of the
Amazon river(figure 6).In the group of very high exposure (1,114 counties), most counties
exposed to landslides are located in the states of Paraná (282), Santa Catarina (165) and São
Paulo (192) and Rio Grande do Sul (271). Of the 20 most exposed counties (table 4), Rio
Grande do Sul is the state with the highest number (7 counties) and the most exposed county
is Lajeado (Tocantins). In this group, 12 counties are located in the Southern Brazil.

Figure 4: Exposure to natural hazards used in DRIB Index (by county).

‘

Figure 5: Exposure to sea level rise in DRIB Index (by county).

 ‘

Figure 6: Exposure map (by county).

Table 4 - Top 20 Counties with highest exposure (values in population)

County State Landslide Flood Droughts SeaLeve Exposure Exposure Population
l Total Index 2010 IBGE
Rise
LAJEADO TO 42855 2820 942 0 46617 2773
1,000
IBIRACU ES 147075 0 3197 0 150272 11178
0,980
JOAO NEIVA ES 197910 0 4417 0 202327 15809
0,976
FIGUEIRAO MS 34590 312 67 0 34969 2928
0,971
QUATRO IRMAOS RS 17820 352 631 0 18803 1775
0,959
IVORA RS 48585 0 367 0 18952 2156
0,943
SONORA MS 122930 802 709 0 124441 14833
0,940
FAXINALZINHO RS 21060 176 334 0 21570 2567
0,939
SANTA RITA DO GO 50540 6416 683 0 57639 6924
ARAGUAIA 0,939
ALTAMIRA DO PR 33330 2028 477 0 35835 4306
PARANA 0,939
DONA FRANCISCA RS 12080 15773 408 0 28261 3401
0,939
 ‘

SANTA ES 90565 7606 3623 555 102349 12240

LEOPOLDINA 0,938
VARGEAO SC 26860 140 544 0 27544 3532
0,933
MARQUES DE RS 22220 8502 490 0 31212 4068
SOUZA 0,931
IGUATU PR 16505 114 327 0 16946 2234
0,930
SANTA FE DE GO 33710 324 751 0 34785 4762
GOIAS 0,927
UNIAO DO OESTE SC 20035 380 429 0 20844 2910
0,925
BARRA DO RIO RS 13795 194 265 0 14254 2003
AZUL 0,925
BARRA BONITA SC 11900 1080 269 0 13249 1878
0,924
PAULO BENTO RS 15045 0 298 0 15343 2196
0,923

In absolute terms, the largest urban areas in Brazil have huge populations
exposed to natural hazards, particularly the cities of Rio de Janeiro (more than 2 million people
exposed to landslides); São Paulo and Porto Alegre (respectively, 3.6 million and more than 3
million people exposed to floods); Sao Paulo, Rio de Janeiro and Fortaleza (respectively, 1.12
million, almost 800,000 and more than 650,000 people exposed to drought).With regard to the
consequences of climate change, major urban areas in Brazil have large populations potentially
exposed to rising sea levels, as shown in the table 5. The cities of Vila Velha and Vitória
(Espírito Santo), Salvador (Bahia) and Santos (São Paulo) present high exposure to sea level
rising, in both absolute and relative terms.

Table 5 – Most exposed cities to sea level rise by absolute population.

County State Population % of total
exposed population
VILA VELHA ES 177,428 0,4278
SALVADOR BA 175,526 0,0710
VITORIA ES 90,059 0,2488
SANTOS SP 81,776 0,1495

3.2 Susceptibility

There is a distinctive division north-south division with regard to

susceptibility in Brazil. Most counties in the North Region (253 counties, 56% of all counties in
the region) have very high levels of susceptibility. Another area with very high levels of
susceptibility is the Northeast Region, with 791 counties (44.1% of all counties in the region).
Thus, 93.71% of the counties of very high susceptibility group are concentrated in the North
and Northeast Regions (figure 7).In the Northeast Region, the state that showed most counties
with very high susceptibility levels is the Maranhão (86.18% of all counties in this state). Of the
20 most susceptible counties (table 6), most are located in the state of Amazonas (15 counties)
and the most susceptible county is Santo Antonio do Ica (Amazonas).
‘

Figure 7: Susceptibility map (by county).

Table 6 - 20 Most susceptible counties

County State Susceptibility
SANTO ANTONIO DO ICA AM 0,887
AFUA PA 0,846
VITORIA DO JARI AP 0,835
IPIXUNA AM 0,827
GUAJARA AM 0,825
AMATURA AM 0,823
TONANTINS AM 0,815
BENJAMIN CONSTANT AM 0,809
EIRUNEPE AM 0,796
IRANDUBA AM 0,784
BOCA DO ACRE AM 0,778
CARAUARI AM 0,777
CUTIAS AP 0,774
BARREIRINHA AM 0,774
LABREA AM 0,771
ENVIRA AM 0,770
ANORI AM 0,768
PARINTINS AM 0,759
LARANJAL DO JARI AP 0,757
ARACOIABA PE 0,756
‘

Specifically, one of the indicators that has contributed to the high levels of
susceptibility in the North and Northeast Regions is the serious lack of access to water supply
and sewage. In the Northeast Region only capital cities and some medium-sized cities have
positive indicators for this variable.On the other hand, virtuallyonlycapitalcitiesand medium-
sizedcitiescontain slums. The indicators for people in households with inadequate wall
materials also follow the general pattern of susceptibility, which concentrates at high levels in
the North and Northeast Regions. Regarding the degree of urbanisation, the highest rates are
concentrated in the Southeast Region, especially in the states of São Paulo and Rio de Janeiro.
Moreover, indicators of dependency ratio, vulnerability to poverty, per capita income, and Gini
index also showed similar spatial patterns to the general pattern of the susceptibility map in
the country: differential pattern north-south.

3.3 Lack of Coping Capacities

Among Brazil’s 5,565counties, 1,114(20.02%) present a high level of lack of

coping capacities, which simply means that 1 in 5 countieshasseriousweaknesseswith regardto
the ability toreact immediatelyorto manageimpacts of adisaster.Overall, there is no apparent
spatial pattern (figure 8), as there was with susceptibility, but the lack of coping capacities in
Brazil is rather widespread. Counties in the states of Minas Gerais (6 counties), São Paulo (4
counties) and Maranhão (3 counties - Brejo de Areia is the county that had the higher level of
lack of coping capacity) are among the 20 higher level counties in terms of lack of capacities for
coping with adverse events or disasters (table 7). Among the 100 counties of higher level of
lack of coping capacities, the result is similar to the previous: 20 counties in São Paulo and 38
in Minas Gerais.
‘

Figure 8: Lack of Coping Capacities map (by county).

Table 7 - Top 20 counties with highest Lack of Coping Capacities

County State Lack of Coping Capacity

BREJO DE AREIA MA 0,885
NOVA IPIXUNA PA 0,846
SAO JOAO DO PARAISO MA 0,839
ENGENHEIRO CALDAS MG 0,824
SAO ROBERTO MA 0,819
SAO SEBASTIAO DA BELA VISTA MG 0,817
JARINU SP 0,817
PONTA DE PEDRAS PA 0,816
SAO JOSE DO BARREIRO SP 0,814
SANTANA DO RIACHO MG 0,812
CUITEGI PB 0,811
VARGEM GRANDE PAULISTA SP 0,805
BARROQUINHA CE 0,804
CAMPESTRE AL 0,802
PAULA CANDIDO MG 0,802
ANALANDIA SP 0,801
DUQUE DE CAXIAS RJ 0,800
MATHIAS LOBATO MG 0,799
SANTANA DE CATAGUASES MG 0,795
NHAMUNDA AM 0,794
‘

The variables ‘Structural measures to reduce disaster risk’11, ‘Disaster risk

management to floods’, ‘Disaster risk management to landslides’ and ‘Number of physicians
per 1,000 inhabitants’, showed high values for the 100 counties with the greatest lack of
coping capacities, contributing to increase the final values of these counties concerning this
indicator.High levels of corruption and mismanagement contributed to the high level of lack of
coping capacities in the counties of Maranhão, Piauí and Bahia states. In addition, high levels
of lack of ‘Structural measures to reduce disaster risk’ for the states of Amazonas and Pará,
caused their counties to appear in the top 100 counties lacking in coping capacities (20 and 24
counties, respectively). These same states are severely lacking of flood risk management,
including the capital city of the state of Amazonas, Manaus. In contrast, highly exposed areas
at risk of landslides, such as mountainous regions of Minas Gerais, Rio de Janeiro, Santa
Catarina and São Paulo, had severe lack of investments in DRR for landslides. States of the
Northeast Region (mainly BA, PI, MA), north of Minas Gerais and Tocantins contained
numerous counties that have notably neglected investments in habitation programmes for
populations vulnerable to disaster risks. The lack of physicians is widespread in the North and
Northeast Regions, while the lack of hospital beds is severe throughout the country, especially
the states of Rio Grande do Norte (RN), Tocantins (TO), Pará (PA) and Rondônia (RO).

3.4 Lack of Adaptive Capacities

Lack of adaptive capacities hotspots can be clearly seen in the states of

Piauí and Maranhão; counties that are part of the arc of deforestation in the Amazon
rainforest; agreste region of the Northeast Region (figure 9).Of the 100 counties with the most
severe lack of adaptive capacities, 75 are located in the Northeast Region of the country,
including 22 counties in Maranhão and 22 in Piauí. Among the top 20 counties in this category,
it is possible once again to highlight the states of Maranhão (4 counties) and Piauí (4 counties)
(table 8).

11
‘Structural measures to reduce disaster risk’ consist of information on the existence of structural
measures to reduce the risk of disasters by floods and landslides: upstream dams to equalise flooding,
construction of macrodrainage channels, park construction, construction of reservoirs for flood
damping, desilting of water bodies, rectification of rivers, increased rail or diversion of water courses,
relocation of the population in risk areas, revegetation, revitalisation of rivers or basins, slope protection
and stabilisation, restoration slopes of embankments, and containment works.
‘

Figure 9: Lack of Adaptive Capacities map (by county).

Table 8 - Top 20 counties with highest Lack of Adaptive Capacities

County State Lack of Adaptive Capacities

BREJO DE AREIA MA 0,829
FRANCINÓPOLIS PI 0,809
JOCA MARQUES PI 0,802
PAULO JACINTO AL 0,794
LUZILÂNDIA PI 0,794
SÃO DOMINGOS DO MARANHÃO MA 0,785
PLÁCIDO DE CASTRO AC 0,777
CAMPESTRE DO MARANHÃO MA 0,777
POÇÃO PE 0,773
TUPIRAMA TO 0,769
MOGEIRO PB 0,766
TOMAR DO GERU SE 0,766
BERIZAL MG 0,765
BARRA DE GUABIRABA PE 0,765
SÃO FRANCISCO DO BREJÃO MA 0,764
RIACHÃO PB 0,762
MASCOTE BA 0,762
SALGADO FILHO PR 0,762
LAGOA DO SÍTIO PI 0,759
‘

SÃO MIGUEL DE TAIPU PB 0,757

Notably, counties in Amazonas and Pará had better results than in

susceptibility and coping capacities indicators due to more favourable values in ‘education and
research’, especially in relation to the percentage of people in primary and secondary
education, as well as good results in environmental policies, deforestation rate (except the
eastern state of Pará), conservation areas and planning tools.The states of MA and PI also had
good results in ‘education and research’, however they had negative results in virtually all
other variables, especially illiteracy rate, students in tertiary education, specific policies for
women, environmental policies, deforestation rate (especially Maranhão), conservation areas,
fire spots (Maranhão) and life expectancy at birth.Furthermore, specific policies for women are
generally neglected by 5,336 of the 5,565 Brazilian counties, that is 99.4%. In addition, most of
the worst results regarding the variable ‘female householder without complete primary
education and with children under 15 years’, are concentrated in the North and Northeast
Regions.With respect to specific policies for disaster prevention, there is no spatial pattern, as
all Brazilian states contain several counties with hardly any such policy. Of the 5,565 counties,
1,689 of them (30.35%) hold little or no legislation or planning tools for preventing disasters12.

3.5 Vulnerability

The hotspots of vulnerability in Brazil (figure 10) are clearly located in the
counties in North and Northeast Regions. Other states hold counties with very high
vulnerability but they are more spatially isolated (Northern Minas Gerais and in the Ribeira
Valley in the south of SP).In the group of counties with highest vulnerability (1,113 counties), 8
states contained 778 counties (69.9% of this group), almost all of them in North and Northeast
Regions, as shown in the table 9. Besides that, in this group, 711 counties are in the Northeast
Region and 182 in the North Region, totalling 893 counties in very high level of vulnerability,
which corresponds to 80.23% of the group and 16.04% of all counties. Among the 100 most
vulnerable counties, 5 states concentrated 69 counties, of which 58 are located in the North
and Northeast Regions (table 10).

12
The indicator explains the presence of local legislation and planning instruments that include specific
planning tools to prevent disasters: master plans including the prevention of flooding or gradual floods
or mudslides or flash floods; laws of use and occupancy including the prevention of flooding, gradual
floods, mudslides or flash floods; specific laws addressing the prevention of flooding, gradual floods,
mudslides or flash floods; master plans including the prevention of landslides or landslides; laws of use
and occupancy including the prevention of landslides or landslides; specific laws that address the
prevention of slips or landslides; county risk reduction plans; geotechnical maps of eligibility for
urbanisation; sanitation plans concerning the service water supply, service sanitation, service street
cleaning and solid waste management, and service drainage and urban stormwater.
‘

Figure 10: Vulnerability map (by county).

Table 9: Group of eight Brazilian states that focus most counties on very high vulnerability
condition (778 counties), by states and regions.
States Number of counties in very Total counties / % Total counties Region
high vulnerability condition state
Alagoas 62 102 60,78 Northeast
Amazonas 53 62 85,48 North
Bahia 129 417 30,93 Northeast
Ceará 59 184 32,06 Northeast
Maranhao 134 217 61,75 Northeast
Minas Gerais 156 853 18,3 Southeast
Pará 67 143 46,85 North
Piauí 118 224 52,67 Northeast

Table 10: A hundred most vulnerable counties in Brazil, by states and regions.
States Number of counties Region
Amazonas 19 North
Pará 13 North
Maranhao 17 Northeast
Alagoas 9 Northeast
Minas Gerais (north) 11 Southeast
‘

Among the top 20 (table 11) most vulnerability counties, it is noteworthy

that 15 are concentrated in the states of Maranhão (6 counties), Amazonas (6 counties) and
Pará (3 counties) alone. Furthermore, among the five most vulnerable counties, two are
located in Maranhão state and two in Amazonas.In the group of counties with lowest
vulnerability (1,113 counties), 85.35% are located in the states of the South Region (Rio G. do
Sul, Paraná, Santa Catarina) and the Southeast Region (São Paulo and Minas Gerais - south).
They hold 246, 165, 141, 262 and 136 counties, respectively, in the low vulnerability group.
Notably, the state of Minas Gerais is one of the states with the highest number of both high
and low vulnerability counties. Also in the lowest vulnerability group are the most populous
state capital cities, including São Paulo, Rio de Janeiro, Brasília (administrative capital of the
country), Fortaleza, Belo Horizonte and cities located in major metropolitan regions, such as
Campinas (SP), São Gonçalo (RJ) and Santo André (SP).

Table 11 - Top 20 most vulnerable counties in Brazil

County State Susceptibility Coping Adaptive Vulnerability
BREJO DE AREIA MA 0,594 0,885 0,829 0,769
IPIXUNA AM 0,827 0,644 0,641 0,704
JENIPAPO DOS VIEIRAS MA 0,630 0,754 0,725 0,703
VITORIA DO JARI AP 0,835 0,590 0,668 0,697
GUAJARA AM 0,825 0,640 0,605 0,690
BELAGUA MA 0,681 0,732 0,658 0,690
ITAMARATI AM 0,660 0,699 0,686 0,682
SANTO ANTONIO DO ICA AM 0,887 0,704 0,450 0,680
PRAINHA PA 0,645 0,724 0,657 0,675
TONANTINS AM 0,815 0,652 0,552 0,673
LAGOA GRANDE DO
MARANHAO MA 0,617 0,746 0,656 0,673
OUREM PA 0,590 0,701 0,720 0,670
CUITEGI PB 0,509 0,811 0,689 0,670
ENVIRA AM 0,770 0,618 0,615 0,668
PINDOBA AL 0,538 0,768 0,696 0,667
CENTRO DO GUILHERME MA 0,608 0,672 0,719 0,666
RIACHAO DO BACAMARTE PB 0,555 0,717 0,722 0,665
NOVA IPIXUNA PA 0,576 0,846 0,571 0,664
DOM JOAQUIM MG 0,517 0,737 0,739 0,664
MARAJA DO SENA MA 0,633 0,630 0,728 0,664

3.6 The DRIB Index

The final result of the construction of the DRIB Index is the Disaster Risk
Map, which is a product of exposure and vulnerability analysis and shows the risk outlook for
the 5,565 Brazilian counties. The map (figure 11) shows the result of the formula shown in the
figure 3, combining/overlaying exposure to natural hazards and climatic changes with
vulnerability in the counties.In general, the strong influence of exposure in the final risk level
can be observed, because the spatial patterns of counties at high and very high risk follow the
general patterns of exposure.
‘

Figure 11: DRIB Index map (by county).

Among the group of counties that are most at risk (1,113 counties), 489
counties (43.9% of this group) are concentrated within 5 states: the state of Paraná contains
the largest absolute number of counties at very high risk of disaster(135 counties, 33,8% of the
counties in its territory); next, the state of Minas Gerais contains 124 very high risk counties
(14.5% of all counties in its territory); São Paulo contains 103counties at very high risk (15.9%
of all counties in its territory); Maranhão has 81 counties and Amazonas has 55 counties at
very high risk (respectively 37.3% and 88.7% of all counties in their territories). In the top 100
of counties at very high disaster risk, once again the states of Maranhão and Amazonas are
heavily represented, containing 15 and 18 counties, respectively, totalling more than one third
of counties in this group. Both in the top 20 and in the top 10 (table 12), the state of Amazonas
is in the group containing the highest number of counties at very high disaster risk in Brazil. In
the top 20 and top 10: Amazonas contains3 and 4 counties, respectively.
 ‘

Table 12 - Top 20 counties for DRIB INDEX

Population
County State Region 2010 Susceptibility Coping Adaptative Vulnerability Exposure Risk

VITORIA DO JARI AP N 13724 0,835 0,549 0,668 0,684 0,812 0,556

N
BOCA DO ACRE AM 32792 0,778 0,583 0,643 0,668 0,812 0,543
N
ITAMARATI AM 8232 0,660 0,618 0,686 0,655 0,824 0,540

MADEIRO PI NE 8034 0,557 0,600 0,705 0,621 0,860 0,534

PEDRAS DE FOGO PB NE 28056 0,539 0,679 0,689 0,636 0,836 0,532

ANORI AM N 18351 0,768 0,625 0,663 0,685 0,776 0,531

N
PRAINHA PA 29342 0,645 0,684 0,657 0,662 0,802 0,531
N
AVEIRO PA 15959 0,652 0,828 0,587 0,689 0,766 0,528

ARAME MA NE 31867 0,628 0,839 0,580 0,682 0,773 0,527

IPIXUNA AM N 25362 0,827 0,563 0,641 0,677 0,777 0,526

NOVA CAMPINA SP SE 9100 0,436 0,792 0,589 0,606 0,868 0,526

LAJEADO TO N 2956 0,504 0,538 0,531 0,524 1,000 0,524

N
CARAUARI AM 27405 0,777 0,503 0,629 0,636 0,822 0,523
N
SANTO ANTONIO DO ICA AM 24327 0,887 0,663 0,450 0,667 0,783 0,522

FRANCINOPOLIS PI NE 5270 0,507 0,503 0,809 0,606 0,861 0,522

NE
MARAJA DO SENA MA 7721 0,633 0,550 0,728 0,637 0,819 0,521
NE
JOSELANDIA MA 15755 0,583 0,722 0,667 0,657 0,789 0,519

ENVIRA AM N 18051 0,770 0,538 0,615 0,641 0,809 0,519

SE
PEDRA AZUL MG 24612 0,499 0,659 0,574 0,578 0,892 0,515
N
JURUTI PA 51483 0,616 0,822 0,595 0,678 0,760 0,515

In the Top 20, 19 counties have very high vulnerability and 8 have very high
exposure, being that these conditions exhibit mainly counties in the North (13) e Northeast (6)
regions. Because of the very high vulnerability in the North and Northeast counties, events
with same magnitude could cause more serious and these counties would have more
difficulties in coping and recover from the impacts of a given disaster if compared with
counties from the south-centre area of the country (figure 12). On the graphic,thisNorth-
Southdifferentiationin terms of riskismost evident.Note that counties risk level are divided
intothose who holdhighexposurebutmoderate vulnerability(counties in the southern
region)andthose who holdhigh vulnerability(counties of the Northernregion)butthat
holdlessexposurethancountiesin the south.
‘

Figure 12: Scatterplot showing on y-axis the values for exposure and on x-axis the values for
vulnerability. In the red circles are represented counties with very high exposure located in the
Center-South of Brazil. In the blue circle, counties with very high vulnerability located in the
North-Northeast of the country (The counties are represented in their respective risk level
classes).

4. Opportunities and limitations

In addition to the intrinsic limitations to the methodology used in this study

(limitation analysis of the methodology adapted in this study is available in Welle et al., 2012
and 2013), this section highlights the main limitations and opportunities to advancing on the
issue of disaster risk indicators, especially in Brazil. Methodological uncertainty in calculating
exposure is due to gaps in the landslides shapefile (triggered by precipitation; available in
PREVIEW Global Risk Data Platform), mainly in the Southeast Region of Brazil. Taking into
account that this region is one of the most landslide prone, these voids can cause a failure in
the calculation of exposure to some counties.
The level of abstraction and generalisation is substantial, taking into
account that each county holds environmental and social specificities the spatial scale of which
is not accounted for in this index’s analysis. This requires further research on a local scale.
Moreover, there are vast differences and inequalities between Brazilian counties. From
territorial point of view alone, some counties hold larger areas than many countries do, but
they are extremely sparsely populated. Furthermore, there are counties with higher
populations than those of many countries and there are counties with greatly reduced
population and territorial areas.
In accordance with the operational limitations (available and reliable data)
it was not possible to extend this disaster risk index to all Brazilian counties - there are
currently 5,570 counties since five new counties (Pescaria Brava and Balneário Rincão, SC;
Mojuí dos Campos, PA; Pinto Bandeira, RS; Paraíso das Águas, MS) were created in 2013, and
the most accessible and used data in the research is date prior to that year.13

13
This research was carried out across 5,565 Brazilian counties.
‘

Limitations due to the available data also precluded more reliable results
with regard to exposure calculation for Fernando de Noronha County. Since there is no
available data on the PREVIEW Global Risk Data Platform database for the Island of Fernando
de Noronha, it was decided to extrapolate exposure to droughts for the entire population of
the island. Exposure to landslides and floods were not calculated, and the exposure calculation
in a climate change scenario of sea level rising was performed in the same way as for the rest
of Brazil. Another limitation is related to the lack of accessible data for the same year, which
complicates the risk analysis for specific years and its evolution in time (the data from the
Human Development Atlas of Brazil and the Profile of Brazilian Counties (Perfil dos Municípios
Brasileiros - MUNIC) are an exception). Also regarding data availability, it should be noted that
since Brazil is heavily exposed to tropical storms, especially the states of the South region (see
Brazilian Atlas Data Disaster - UFSC.CEPED, 2013), it is essential that there is a systematisation
of geo-referenced data on cyclones and tropical storms trajectories in Brazil, which would
allow the calculation of the population exposed to these phenomena, as is the case with
studies by Peduzzi et al. 2012.
For this research, it was not possible calculate exposure to tropical storms
in Brazil since there is no systematic data collection for cyclone displacement trajectories in
the country. NOAA (National Oceanic and Atmospheric Administration) conduct a
comprehensive survey on the subject, however it considers only so-called tropical cyclones
(hurricanes / typhoons). As such, only two events of tropical cyclones have been recorded in
Brazil, whose trajectories can be viewed in the figures available from the NOAA websites
(explanation given in an interview by Daniel Henrique Candido, Operational Analyst in Natural
Disasters of the National Center for Natural Disaster Monitoring and Alerts - CEMADEM, Brazil,
2014).14
Finally, it is necessary to highlight a potential limitation of this index with
regard to the data from the Profile of Brazilian Counties (MUNIC). Since these data come from
information provided through questionnaires answered by the cities’ own managers (or their
representatives), and not through consulting official documents, there is the potential risk of
introduction of incorrect information provided by the public administrator as to the actual
disaster risk management situation and other issues in the county (despite the arguments
present in MUNIC report (IBGE, 2014), that all precautions to avoid incorrect information in
the survey were taken out).15

5. Discussion

14
An attempt at calculating exposure to tropical storms was performed with data kindly provided by Dr.
Claudine P. Dereczynski and Ronaldo Palmeira, UFRJ, Brazil, used in the report ‘Extratropical Cyclones
Occurrence Probability in Campos and Santos’. However, this data treated tropical storm trajectories
not in terms of time and space, but rather in terms of pressure of atmospheric, indicating the possibility
of occurrence of tropical storms, but not allowing the calculation of exposure since there was in fact no
validating data (compared to real events).
15
‘At this point, IBGE, upon receiving the Basic Questionnaire filled, made a visual assessment of the
information collected in order to detect problems in filling, so that they were resolved immediately with
the informant. Data entry was done in a decentralised manner for the supervision of research at the
headquarters of each State Unit, or the technician responsible for the collection, in their own agencies.
Criticism of consistency of the data collected, in turn, was held in each unit, but has also made one
calculation work of information by the staff of the Department of Population and Social Indicators,
responsible for Munic (IBGE, 2014)’.
‘

As in Welle et al. (2013), it is not the intention of this research to exhaust

the complexity of hazards, their origin, magnitude, frequency, and possible impacts, and to
scrutinise all the multiple dimensions of vulnerability. However, it is necessary to recognise
that, as in similar studies, not only can natural hazards be attributed to the normal everyday
activities of the people, but also social, cultural, economic, governance and empowerment are
key elements for defining whether a natural hazard can materialise into a disaster. To that
effect, this research provides an adaptation of a statistical model and a substantially coherent
approach to establishing the spatial patterns of risk and vulnerability to natural disasters in
Brazil. Discussions about the robustness, reliability, reasonability, and validity of the
methodology used in this study are available in Welle and Birkmann (2015).
Welle et al. (2013) has specified that since large countries such as Brazil,
China, and India are more likely to contain different regions with different types of natural
hazards, the WorldRiskIndex method is designed with a modular structure, allowing it to adapt
analysis to regional and local scales. Thus, the primary structure of the method, which analyses
exposure, susceptibility, lack of coping and adaptive capacities, allows an important synthesis-
panorama of disaster risk conditions at the local (county) scale in Brazil.
Above all, the results analysis of natural disasters risk index in Brazil shows
that the counties in very high risk situations have very high exposure to natural hazards,
especially counties located in the South Region, but hold relatively moderate to low
vulnerability, making these counties less susceptible, more capable of coping with the
occurrence of these phenomena and adapting to social and environmental changes that may
occur in the medium to long term. On the other hand, counties located in the basin of the
Amazon River (mainly) and in the Northeast Region face high exposure conditions to natural
hazards (not as high as the counties in the South Region), however they have serious
conditions of susceptibility, very low capacities for dealing with, coping, facing and recovering
from adverse conditions that arise when disaster strikes, and have very low capacity for
adapting to current and future social and environmental changes, taking into account climate
change scenarios.
As demonstrated in the presentation of the results, 80.23% of the counties
in very high vulnerability condition are located in the North and Northeast Regions, especially
in the latter. This means that 13.6 million people live in counties that have very high
vulnerability in the Northeast Region alone, out of a total of 24.9 million people (in very high
vulnerability group) living in counties that are most susceptible, have lower capacity for coping
with, recovering from and adapting to changes imposed by the impacts of disasters and
climate change.The clear North-South spatial division in terms of vulnerability patterns is
defined mainly by the huge socio-economic inequalities between these two poles of the
country, historically constructed over the past 500 years. Indicators of susceptibility, the lack of
coping capacities, and the lack of capacity to adapt to changes imposed by disasters and
climate change, can contribute to buttress regional and intra-regional inequalities in Brazil
which expose the Northeast Region as it holds the highest vulnerability rates in the country.In
the case of lack of coping capacity, there is an alarming overlap between, on the one hand,
sets of counties serious lacking in DRR indicators and, on the other, areas/regions that are
heavily exposed and that display recent disasters history. The Northeast Region is also a source
for concern because of its lack of adaptive capacities. Among the counties that are very highly
lacking in adaptive capacities are those located in Maranhão and Piauí, which show negative
indicators in illiteracy rate, mothers as household heads without completed primary education
and with children under 15 years, specific policies and actions for the environment , proportion
of deforestation, life expectancy at birth, fire spots, legislation and planning instruments and
conservation areas.These are counties with low capacity for adapting to changes imposed by
disasters and/or climate change. Maranhão is among the states with the highest deforestation
rates in the Amazon rainforest, low rates of existence of conservation areas, high rates of fire
‘

hotspots, while showing very negative social indicators. These factors combined place
Maranhão among the states most vulnerable to disaster risk and climate change in Brazil.
It is evident that the disaster risk indicators in Brazil exhibit, once again, one
North-South polarity that distinguishes counties exposed to floods and droughts (landslides in
isolated spaces) in the North and Northeast Regions, which hold very high levels of
vulnerability, resulting in high levels of disaster risk due to the high susceptibility, very low
capacity to cope with and adapt to the socio-environmental changes imposed by disasters and
climate changes; and counties greatly exposed to an overlap of multiple hazards (landslides,
floods, flash floods, and droughts). The high population density located in high hypsometry
and steep slopes areas, natural conditions (climate, geological structure and paedological
conditions), makes them more prone to occurrences of mass movements while they are
exposed to floods, droughts and potentially to sea level rise. However, exhibit lower levels of
vulnerability, which makes these counties, especially those in South Region, more capable of
withstanding, recovering, coping with, and adapting to the socio-economic, cultural and
environmental changes that may arise with occurrence of disasters and climate changes.

6. Conclusions

This research represents an effort to assess vulnerability and risk in a more

comprehensive way, considering indicators of exposure, susceptibility, coping and adaptive
capacities. This disaster risk index is a quantitative approach that seeks a concept for risk and
vulnerability analysis integrating national/local scales. This type of approach has great
potential for measuring exposure, susceptibility, coping and adaptive capacities because it
uses indicators that outline revealed vulnerability of past events (Welle and Birkmann, (2015).
Nevertheless, the ability of these tools/datasets to capture local-specific
risk and vulnerability is limited to a certain extent, then it is necessary approaches at the local
level to capture specific characteristics of vulnerability (e.g. social networks, risk governance
and performance of local governments in disaster risk management) and exposure (e.g. geo-
environmental context assessment).
Specifically, the results of disaster risk indicators in Brazil showed that the
risk is strongly interwoven with social-economic and cultural conditions and normal everyday
life, as well as with the performance of state institutions dealing with DRR, in other words,
vulnerability. Spatial trends of disaster risk and vulnerability, products of this research, also
have stressed the serious inequalities between and within regions of the country, which result
in barriers to the development of the DRR in Brazil as a whole.
In addition, the use of the results of this index, both taken in its entirety,
but mainly taken the components individually, has a significant potential to contribute to
decision making and specific actions to DRR. As a whole, the method reveals the spatial
disparities regarding the potential risk of the population to natural hazards. However, if
components are taken in a disaggregated and isolated manner, the results highlight the serious
conditions in terms of social indicators (susceptibility), the fragility on the decision making and
basic structures to deal with disasters in Brazilian municipalities (coping capacities), and
explain problems of environmental order , social and governance with regard to the adaptive
capacities to environmental changes imposed by the disasters and climate changes.
Therefore, the quantitative results and the spatial patterns that have been
established on disaster risk indicators may stimulate further discussions, at the academic level
and at the level of risk governance, on how to reduce exposure and susceptibility on the one
hand, and on the other hand increase/improve the capacities to cope and adapt to the
consequences of natural hazards. Despite recent improvements related to planning, decision-
making, risk culture and academic discussion of disasters in Brazil, especially after the disaster
in the mountainous region of Rio de Janeiro in January 2011, much effort is still needed to
‘

make the country more resilient to current/daily and future disasters, with special attention to
the local level, with the county being the most problematic,16 from a DRR point of view.
Considering the expanse of Brazilian territory, the shortcomings with regard
to DRR decision-making, in terms of research and production of knowledge on the subject
(especially at the local level), and given that the subject is trendy and recently there is an
increased awareness of the issue in Brazilian society (governments, business, NGOs and society
as a whole), the results of this study, with the overview and spatial patterns of disaster risk in
Brazil, has a huge potential to establish further research at the local level (counties,
metropolitan regions, river basins, etc.), also with great possibilities for inter-institutional
partnerships within Brazil, and between Brazilian and foreign research institutions, adding
researchers from multiple disciplines to improve the performance of Brazil's capacity for DRR
through knowledge production.

7. Appendices

Appendix A

A1: Composition of the susceptibility component.

16
Regarding the challenges of municipal management in Brazil, see Velloso et al. (2011).
‘

A2: Composition of the coping capacity component.

A3: Composition of the adaptive capacity component.

8. Acknowledgements
The first author would like to thank the Institute for Environment and Human Security of
United Nations University (UNU-EHS) and the Institute of Regional Development Planning of
University of Stuttgart for awarding a visiting scientist position to carry out research. The
research leading to these results has received funding from the Coordenação de
Aperfeiçoamento de Pessoal de Nível Superior (CAPES) under Grant Agreement no. 4289/14-5
(Post-doctoral scholarship). We are also grateful to Leila Sousa and Cristiano Alves forvaluable
support on the methods and GIS.

References

1. Almeida, L. Q.(2010) - Vulnerabilidades socioambientais de rios urbanos. Bacia

hidrográfica do rio Maranguapinho. Região Metropolitana de Fortaleza, Ceará. Post-
Graduate in Geography. Instituto de Geociências e Ciências Exatas/Unesp, Rio Claro
(São Paulo, Brazil), Doctoral Thesis, 278p. Available in
http://www.athena.biblioteca.unesp.br/exlibris/bd/brc/33004137004P0/2010/almeid
a_lq_dr_rcla.pd
2. Alves, Humberto Prates da Fonseca, 2006. Vulnerabilidade socioambiental na
metrópole paulistana: uma análise sociodemográfica das situações de sobreposição
espacial de problemas e riscos sociais e ambientais. Rev. Bras. Estudos. Populacionais.
vol.23 nº.1 São Paulo Jan./Junho 2006.
‘

www.scielo.br/pdf/rbepop/v23n1/v23n1a04.pdf
3. Barry, E. (2015, May 05). Nepal’s Young Men, Lost to Migration, Then a Quake.
International New York Times.
http://www.nytimes.com/2015/05/05/world/asia/nepal-already-stripped-of-young-
men-faces-a-darker-problem.html
4. Billing P.; Madengruber U. (2005) Coping Capacity: towards overcoming the black hole.
European Commission: Directorate-General for Humanitarian Aid (ECHO). World
Conference on Disaster Reduction, Kobe / Japan, 18-22 Jan 2005.
http://ec.europa.eu/echo/files/evaluation/watsan2005/annex_files/ECHO/ECHO4%20
-%20coping%20capacity%20article%20-%20final.doc
5. Birkmann, J. (2013) Measuring vulnerability to promote disaster-resilient societies:
conceptual frameworks and definitions’, in Birkmann, J. (Ed.): Measuring Vulnerability
to NaturalHazards: Towards Disaster Resilient Societies, 2nd ed., pp.9–80, United
Nations University, New York.
http://unu.edu/publications/books/measuring-vulnerability-to-natural-hazards-
towards-disaster-resilient-societies-second-edition.html#overview
6. Birkmann, J., 2006: Measuring Vulnerability to Natural Hazards – Towards Disaster
Resilient Societies. United Nations University Press, Tokyo, Japan, 450
pp.archive.unu.edu/unupress/sample-chapters/1135-
MeasuringVulnerabilityToNaturalHazards.pdf
7. Birkmann, J., Welle, T., Krause, D., Wolfertz, J., Catalina Suarez, D. and Neysa Setiadi
(2011): WorldRiskIndex: Concept and Results. In: WorldRiskReport 2011. Alliance
Development Works, 13-42, ISBN 978-3-9814495-1-8.
http://www.worldriskreport.org/
8. Bogardi, J. and J. Birkmann, (2004) Vulnerability assessment: the first step towards
sustainable risk reduction. In: Disasters and Society – From Hazard Assessment to Risk
Reduction [Malzahn, D. and T. Plapp (eds.)]. Logos Verlag, Berlin, Germany, pp. 75-
82.collections.unu.edu/view/UNU:2808
9. Brasil. Sistema Nacional de Informações sobre Recursos Hídricos (SNIRH). 2015.
Vulnerability Map to Floods – Brazil. Available in:
http://www2.snirh.gov.br/home/index.html
10. Cannon, T., (1994) Vulnerability analysis and the explanation of ‘natural’ disasters. In:
Disasters, Development and Environment [Varley, A. (ed.)]. John Wiley and Sons,
Chichester, UK, pp. 13-29.
http://trove.nla.gov.au/work/6990102?selectedversion=NBD10812233
11. Cardona, O D (1999) Environmental management and disaster prevention: Holistic risk
assessment and management, in J Ingleton (ed) Natural Disaster Management, Tudor
Rose, London.
12. Cardona, O.D., (1986)Estudios de vulnerabilidad y evaluación del riesgo sísmico:
Planificación física y urbana en áreas propensas. Boletín Técnico de la Asociación
Colombiana de Ingeniería Sísmica, 33(2), 32-65.
13. Cardona, O.D., M.K. van Aalst, J. Birkmann, M. Fordham, G. McGregor, R. Perez, R.S. et
al. (2012) Determinants of risk: exposure and vulnerability. In: Managing the Risks of
Extreme Events and Disasters to Advance Climate Change Adaptation [Field, C.B., V.
Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, et al. (eds.)]. A Special Report of
Working Groups I and II of the Intergovernmental Panel on Climate Change (IPCC).
Cambridge University Press, Cambridge, UK, and New York, NY, USA,pp. 65-108.
http://www.ipcc-wg2.gov/SREX/
14. Costa, M. A.; Marguti, B. O. Atlas da vulnerabilidade social nos municípios brasileiros.
Brasília : IPEA, 2015.ivs.ipea.gov.br/ivs/data/rawData/publicacao_atlas_ivs.pdf
15. Cutter, S. L. ;Boruff, B. J. ; Shirley, W. L. Social vulnerability to environmental hazards.
Social Science Quarterly, 84 (1):242-261, 2003.
‘

http://onlinelibrary.wiley.com/doi/10.1111/1540-6237.8402002/full
16. Dauphiné, A. (2005) - Risques et catastrophes. Observer, spatialiser, comprendre,
gérer. Paris: Armand Colin. http://www.armand-colin.com/risques-et-catastrophes-
9782200278427
17. Deschamps, M. V. Vulnerabilidade socioambiental na região metropolitana de
Curitiba/PR. Curitiba, Tese (Doutorado em Meio Ambiente e Desenvolvimento) –
Universidade Federal do Paraná. 2004.
155p.www.ipardes.gov.br/biblioteca/docs/tese_marley_deschamps.pdf
18. Hewitt, K. Regions of risk: a geographical introduction to disasters. Longman, 1997.
19. IBGE. Perfil dos municípios brasileiros 2013. Rio de Janeiro: IBGE, 2014.
http://www.ibge.gov.br/home/estatistica/economia/perfilmunic/2013/default.shtm
20. Instituto de Estudios Ambientales (IDEA) (2005) Indicators of Disaster Risk and Risk
Management: Main Technical Report, IADB/IDEA Program of Indicators for Disaster
Risk Management, Universidad Nacional de Colombia,
Manizales.www.iadb.org/document.cfm?id=35177671
21. International Federation of the Red Cross and Red Crescent Societies (2008) Guidelines
for Assessment in Emergencies [online]
http://www.icrc.org/eng/assets/files/publications/icrc-002-118009.pdf (accessed 30
may 2015).
22. IPCC, 2012: Managing the Risks of Extreme Events and Disasters to Advance Climate
Change Adaptation. A Special Report of Working Groups I and II of the
Intergovernmental Panel on Climate Change [Field, C.B., V. Barros, T.F. Stocker, D. Qin,
D.J. Dokken, K.L. Ebi, et al. (eds.)]. Cambridge University Press, Cambridge, UK, and
New York, NY, USA, 582 pp.http://www.ipcc-wg2.gov/SREX/
23. Lewis, S. L.; Brando, P. M.; Phillips, O. L.; Heijden, G. M. F.; Nepstad, D. The 2010
Amazon Drought. 4 FEBRUARY 2011 VOL 331 SCIENCE www.sciencemag.org.
Downloaded from www.sciencemag.org on April 29, 2015.
24. Maskrey, A. (1993) Los Desastres No son Naturales. Red de Estudios Sociales em
Prevención de Desastres en América Latina, LA RED, Tercer Mundo Editores, La RED,
Bogotá, Colombia.www.oei.es/decada/portadas/Desnat.pdf
25. Preview: Global Risk Data Platform. Database. Available in:
http://preview.grid.unep.ch (access in: 28.09.2014).
26. Peduzzi, P., Dao, H., Herold, C. and F. Mouton, 2009: Assessing global exposure and
vulnerability towards natural hazards: the Disaster Risk Index. In: Natural Hazards and
Earth System Sciences 9, S. 1149-1159. nat-hazards-earth-syst-
sci.net/9/1149/2009/nhess-9-1149-2009.pdf
27. Thywissen, K., 2006: Core terminology of disaster risk reduction: A comparative
glossary. In: Measuring Vulnerability to Natural Hazards [Birkmann, J. (ed.)].UNU Press,
Tokyo, Japan, pp. 448-
496.www.eng.uwo.ca/research/iclr/fids/publications/conferences/Dec2004/Comparat
iveGlossary_V2.pdf
28. UFSC.CEPED. Atlas Brasileiro de Desastres Naturais: 1991 a 2012. 2nd ed. rev..ampl.
Florianópolis: CEPED/UFSC, 2013. http://150.162.127.14:8080/atlas/atlas.html
29. UNDP, 2004: Reducing Disaster Risk: A Challenge for Development, A Global Report.
UNDP, New York, NY.www.preventionweb.net/files/1096_rdrenglish.pdf
30. UNDRO, 1980: Natural Disasters and Vulnerability Analysis. Report of Experts Group
Meeting of 9-12 July 1979, UNDRO, Geneva,
Switzerland.https://archive.org/details/naturaldisasters00offi
31. UNISDR (2015). Sendai Framework for Disaster Risk Reduction 2015-2030. Third United
Nations World Conference on DisasterRisk Reduction. Sendai, Japan, 14-18 March
2015. Available in:
‘

http://www.wcdrr.org/uploads/Sendai_Framework_for_Disaster_Risk_Reduction_201
5-2030.pdf
32. UNISDR, 2009: Global Assessment Report on Disaster Risk Reduction: Risk and Poverty
in a changing climate-invest today for a safer tomorrow, Geneva, 207pp.,
http://www.preventionweb.net/english/hyogo/gar/report/index.php?id=9413
33. UNISDR. (2004) Living with Risk: A global review of disaster reduction initiatives.
Volume I, New York and Geneva: United Nations.www.unisdr.org/files/657_lwr1.pdf
34. UNISDR (2009) Terminology on Disaster Risk Reduction. UNISDR,
2009.http://www.unisdr.org/we/inform/publications/7817
35. Veloso, J. F. A. et al. (2011). Gestão municipal no Brasil: um retrato das prefeituras.
Brasília: Ipea, 303 p. Available at: http://www.cepam.org/media/143883/Ipea_Livro-
Gest%C3%A3o-Municipal-no-Brasil.pdf [accessed 24 June 2015].
36. Welle, T., Birkmann, J., Krause, D., Suarez, D.C., Setiadi, N. and Wolfertz, J. (2013) The
WorldRiskIndex: a concept for the assessment of risk and vulnerability at
global/national scales, in Birkmann, J. (Ed.): Measuring Vulnerability to Natural
Hazards: Towards Disaster Resilient Societies, 2nd ed., pp.219–251, United Nations
University, New York.http://unu.edu/publications/books/measuring-vulnerability-to-
natural-hazards-towards-disaster-resilient-societies-second-edition.html#overview
37. Welle, T., &Birkmann, J. (2015): The World Risk Index – An approach to assess risk and
vulnerability on a global scale. In: Journal of Extreme Events, Vol. 2, No. 1, (34
pages)http://www.worldscientific.com/doi/abs/10.1142/S2345737615500037
38. Welle, T., Birkmann, J., Rhyner, J., Witting, M. and Wolfertz, J. (2012) WorldRiskIndex
2012: concept, updating and results, WorldRiskReport 2012, pp.11–27, HRSG, Bündnis
Entwicklung Hilft, Aachen, Germany.http://www.worldriskreport.org/
39. WorldRiskReport 2014, HRSG. Bündnis Entwicklung Hilft und UNU, pp. 39-53, ISBN
978-3-9814495- Welle, T., Birkmann, J., Rhyner, J. (2014): The WorldRiskIndex 2014. In:
3-3http://www.worldriskreport.org/
40. Wisner, B.; Blaikie, P.; Cannon, T. & I. Davies (2004): At Risk: Natural Hazards, People´s
Vulnerability and Disasters. (Routledge) London, New
York.www.preventionweb.net/files/670_72351.pdf

Highlights

An adaptation of World Risk Index on the national scale is proposed.

Model tested for Brazil using different databases and indicators.

Disaster risk index model for aggregated indicators of exposure and vulnerability.

Results provide comparison between counties regarding the various disaster risk components.

Results provide spatial patterns of disaster risk by county in Brazil.

‘